Bimetalic - electromechanical - galvanic corrosion

[RMMazzoli]

We're a PCB manufacturer and have requests to build boards that have two or more metal finishes.&nbsp;&nbsp;Many times these finishes are electroplated onto the traces and pads.&nbsp;&nbsp;A problem we encounter is when we use these multiple metals as the &quot;etch resist&quot;.&nbsp;&nbsp;For those who aren't familiar with the process, basically we photo-define traces and pads then plate up copper onto those areas.&nbsp;&nbsp;Underneath the photo-film is the base copper which acts as continuity for the plating process.&nbsp;&nbsp;After copper plating we may plate tin, tin/lead, or gold (or combinations in selected areas).&nbsp;&nbsp;We then strip the film and have to etch the background copper.<br><br>Therein lies the problem.&nbsp;&nbsp;We use an ammonia etchant to etch the background copper.&nbsp;&nbsp;When we use only one &quot;etch resist&quot; then we have no problem.&nbsp;&nbsp;When we use two, say gold and tin/lead, then the copper is etched faster in the areas where these metals meat than anywhere else.&nbsp;&nbsp;The industry defines this as electromechanical corrosion and has no good method for dealing with it.&nbsp;&nbsp;<br><br>The problem is twofold: 1) background copper is etched at a faster pace in bi-metalic areas, and 2) the traces we're trying to protect with the resist are &quot;undercut&quot; due to the enhanced speed of the etchant in those areas.<br><br>Are there any corrosion experts here that can offer some solutions?&nbsp;&nbsp;Like all PCB manufacturers we have band-aids for the problem, but no real solution.&nbsp;&nbsp;Any help will be appreciated.&nbsp;&nbsp;<br><br>Thanks<br><br>Mark Mazzoli<br><A HREF="mailto:MMazzoli@protoeng.com">MMazzoli@protoeng.com</A>